Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/248604 
Year of Publication: 
2021
Series/Report no.: 
ISER Discussion Paper No. 1145
Publisher: 
Osaka University, Institute of Social and Economic Research (ISER), Osaka
Abstract: 
The observability of partners' past play is known to theoretically improve cooperation in an infinitely repeated prisoner's dilemma game under random matching. This paper presents evidence from an incentivized experiment that reputational information per se may not improve cooperation. A structural estimation suggests that a certain percentage of players act according to the "Always Defect" strategy, whether or not the reputational information is available. The remaining players adopt available cooperative strategies: specifically, the tit-for-tat strategy when reputational information is not available, and a strategy that conditions on the matched partner's past play when reputational information is available.
Subjects: 
experiment
cooperation
infinitely repeated prisoner's dilemma game
reputation
JEL: 
C92
C73
D70
Document Type: 
Working Paper

Files in This Item:
File
Size
7.02 MB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.